CS 6501 3D Computer Vision (Fall 2025)

University of Virginia

Course Description


The ability to perceive the world in 3D is critically important for humans and has extensive applications in robotics, graphics, virtual/augmented reality, and more. This course will delve into the foundational concepts and recent advancements at the intersection of machine learning and 3D computer vision. In particular, we will cover the following topics in this course:

  • Classical Multiview Geometry
  • Explicit, Implicit, and Neural 3D Representations
  • Differentiable rendering
  • Single-view 2.5D and 3D Prediction
  • Generative 3D models
  • Modeling 3D in Time
  • 3D Meets Robotics and Beyond

Course Objective


Upon the completion of this course, the students should:

  • have a big picture of the history and recent trends in 3D computer vision;
  • be able to analyze the recent 3D vision papers critically;
  • identify compelling future research questions within the field;
  • learn how to communicate and collaborate in research, and how to present efficiently;

Course Staff


Zezhou Cheng

Instructor: Zezhou Cheng

Email: zc3bp@virginia.edu

Office hour: 5:00-6:15pm Tuesday

Location: Rice hall, Room 502

Lab: Computer Vision Lab @ UVA

Jin Yao

TA: Jin Yao

Email: rry4fg@virginia.edu

Office hour: 10:00-11:15am Friday

Location: Rice 414 Meeting Room; Zoom link

Lab: Computer Vision Lab @ UVA

Course Platforms


  • Ask questions, look for teammates, check out announcements on Piazza
  • Submit assignments, presentation slides, or project reports on Canvas
  • Give us (anonymous) feedbacks on Canvas

Course Format


The course format will include a combination of lectures, student-led presentations, and course projects. To have fun, we adopt the Role-Playing Paper-Reading format in the student-led presentations.

Time and Location


  • Dates: 08/26/2025 - 12/09/2025
  • Meets: MoWe 5:00PM - 6:15PM
  • Room: Rice Hall 340, Main Campus

Acknowledgement


The course material borrows heavily from the following seminar lectures: